As pointed out by dsw in the Daily Paul's post below, a really strange gap is visible between Paul's popular vote in Alabama's primary and thevote count in favor of his delegates, as it all happens on the same day through the same ballot. If you are not familiar with the enigma addressed here, I can only encourage you to read the post linked below first.
I think that it is worthwhile separating the discussion of this one in its own thread as the mathematical analysis and speculative interpretations are quite different from the other issues raised elsewhere.
As per Republican Party Rules, you could only vote for 1 candidate, and then only for the delegates of that candidates. For each of your candidate's delegate proposed, you had to choose between a Mr Jones or a Mr Smith to get the job. A rather complicated affair as illustrated by the typical ballot:
And that is only the first of the two pages...
A Gingrich supporter had to trawl through a list of 9 separate delegate contests, Paul 16, Romney 18 and Santorum 3. Voter's fatigue is very visible in the data: almost 20% of the electorate failed to reach the end of the delegate list after it started to fill it in!
I have collected all the precinct information available as of today (60 counties out of 67), ie 1,864 precincts x 46 delegate ballots = 85,744 data points. I have not included the additional Congessional District Delegate races, only the statewide ones.
The chart below verifies and confirms dsw's observation:
On my data, 82,940 votes were cast for Paul's 1st delegate contest, but only 29,609 for him directly as President choice. -64%.
187 precincts have votes for Paul's 1st delegate, but none for him as President. Yes, that is 10% of them all.
In Lowndes County, votes for Paul as President are only a 10th of the votes for his 1st delegate on the list.
In Walker County, in the "Prospect Method. Chuch" precinct, Paul's 1st delegate contest collected 34 votes. Paul as a choice for President? 1 vote. One vote.
As can be seen on the chart, the other candidates tend to get 1 vote in the delegate contest for 1 in the President's choice, with little volatility above and below that, as per the Republican party rule reiterated on the ballot itself. Paul's distribution of votes is from a different planet.
More to come.
Edit: How unusual is it to have 1/3 of the votes in the presidential preference vote that you get in the delegate races? Here is a partial answer: